AI-Based Conflict Detection and Decision Support in Air Traffic Control: A Critical Review of Emerging Approaches

M2S2I Laboratory, ENSET Mohammedia, University Hassan 2, Casablanca Morocco

^* wafaeaouladhadj@gmail.com

Abstract

The increasing complexity of modern airspace operations necessitates the use of sophisticated technologies that are able to facilitate early conflict detection and effective decision-making in air traffic control. The recent progress in artificial intelligence, particularly machine learning and deep reinforcement learning, has opened new opportunities for conflict detection and resolution enhancement. This paper provides an in-depth analysis of the state-of-the-art in intelligent and adaptive decision-support systems, with emphasis on research conducted between 2020 and 2024. Furthermore, it explores different AI techniques such as supervised learning, behaviour modelling, and multi-agent deep reinforcement learning. Despite strong simulation results, real-world deployment faces challenges such as interpretability and coordination under uncertainty. The review identifies future directions including explainable AI and hybrid models using real-time aviation data. Unlike earlier ATC/AI surveys, this review uniquely integrates conflict-detection and decision-support perspectives within a unified SLR framework, offering a quantitative synthesis across AI paradigms and mapping identified gaps to future research pathways.